Thoughts on Koine Pedagogy

Thought I'd put up some of my meditations on teaching Greek. I've been thinking for quite a while about the fact that we teach ancient languages and modern languages differently. Grammar-Translation vs. Natural Language pedagogy. Obviously, teaching Koine, there's a whole lot of concepts that it's necessary to train into the student -- three sets of declensions, five cases, gender-number-case alignment, verb conjugations, verb tenses, moods, and voices. Not to mention the basics of character and phoneme recognition in a non-Latin alphabet. But the standard approach I see, even in the newer, "glossier" textbooks, is still dumping batches of information on the student and asking them to interpret, but not to use.

I realize that in the seminaries we generally only get one, maybe two terms to teach the language comprehensively -- and the same for Hebrew. And for non-Classics majors in undergrad, two terms is a maximal expectation. And so we cheat. We can either aim for higher usage proficiency or higher concept knowledge, but not both. In other words, full grammar and limited reading competence, or better reading/speaking/listening competence and less grammar. In Bible, this always pushes us toward Grammar-Translation models, where recognition is about all we can achieve in the time given to us. And every New Testament prof is reteaching the basics, come translation time.

What's worse is that when the students get to situations where they need to re-learn all those basics, and their nuances in real texts, teaching nuance has to begin with tearing down the abstractions the student learned first, like "genitive = 'of' = possessive." But that's another tradeoff -- do you teach the full breadth of possible meanings for each case, some restricted set, or some abstraction, and what abstraction is simple enough to do the job and stick with the student? Every gloss fails at some point, but it gives an easy handle.

The less time we have, the more usage we have to cram in to make up for it, because the only thing that earns space between the ears is sustained presence between the ears. And the best way to get that is to keep the language moving in the eyes and ears, and out the mouth and hands.

That's easy enough when you start with phonetics. Basic alphabet, accents and breathings, vowels and diphthongs, consonant classes (P, T, and K). All that stuff teaches orality, if done right. And it needs to be combined with vocabulary -- but the vocabulary needs to come in usable sequencing of its own.

Running through the noun system according to morphology may be conceptually elegant, and usable for a reference grammar, but it seems to me that it doesn't teach usage as well as it teaches recognition. I want students to be able to grab any word and decline it -- eventually. But if I run them through the basic ideas behind each of the cases, and then run them through first, second and third declensions as paradigms, they will spend the most brain-power memorizing the paradigms and not learning the case meanings. In modern languages, we only do that with verb conjugation, where the sole area of change across the paradigm is already indicated by the pronouns. In other words, we only do it when the changes are for person/number agreement, and not meaning. If we do that for verbs, why not do the equivalent for nouns?

Why not teach gender/number paradigms for each case, instead of case paradigms for each declension? Begin with adjectives, because they cover more paradigms at once. Introduce the articles by gender/number paradigms, doing what articles do: making substantives. Then introduce nouns with their articles, which will obviously be of restricted gender. We start with the nominative case, and we teach the production of subject complexes of nouns, articles, and modifiers. In this way, by separating the cases, we also teach them as syntax elements. Build on the genitive case next, and teach modification of subject complexes by quality. (Obviously, we'll do genitives with objective cases when we get to the objective cases, but that will simply re-use what we teach here.)

Now, teach basic verbs -- present active indicative. We're only equipped to do equations and definitions yet, identity statements -- but we are equipped to do them! So eimi, which they have to memorize as irregular anyways, is a given, and we can do predicate nominatives. We teach eimi with the pronouns anyways. We need intransitives now, because we don't have object cases yet. (A problem that doesn't exist in most modern languages! The pronouns change case, but the rest of the noun system doesn't. So object and subject nouns are the same, which means that for Greek I have to come up with a different answer.) But we can't teach object cases until we have started on verbs, so next we teach accusative case and transitive verbs. (Do we teach infinitives and their objects here, or keep waiting until we clear the present indicative? Probably wait.) Again, this lets us expand the range of useful sentences, building the new on top of the old. And last, dative case: indirect objects and adverbial constructions, along with adverbs.

At this point the student has a functioning present tense grammar and a fairly thorough grasp of the noun system, most of which we can take right along into the rest of the verb system. If I had the time, in high-school terms, this would be quite enough for level 1, and with all the extra forms to recognize, it's probably leaking over into the level 2 curriculum. Now, the obvious question is: I've structured this for composition and speaking usage; how well can I bend this sequence to align with translation/reading of genuine passages? How much do I have to bend passages to make them useful in this restricted set? What do I need to add that I haven't got in here? I need to look into graded readers.

Anyhow, moving along: obviously, I haven't done anything with differentiating the active, middle and passive voices. I have to cover aorist, imperfect and future tenses, as well as the perfect. And then we're still in the indicative mood! We need infinitives and participles, and the subjunctive mood at least.

Oh, and particles! And prepositions! Crap, I left those out.

Anyhow, right now, I'm basically only worried about sequence and how well it solves the problem of learning the language well the first time. Two standard programming axioms come into play here: "don't worry about performance" and "premature optimization is the root of all evil." Or, "haste is of the devil." If you start cheating for time before you've written out the algorithm, you will wind up with either a bad solution, or none at all. And I think that's already what we have! We optimize after we have solved the problem the "right" way. So if you're reading this, I know bloody well this is too slow to go about it -- but the question right now is, "can we make it work better," not how much faster it can go.